The 15–90 day delinquency rate on Mercado Pago personal loans for first-time borrowers in Mexico rose from 4.6% to 6.1%. How would you investigate?

Introduction

A rise in 15–90 day delinquency from 4.6% to 6.1% among first-time borrowers on Mercado Pago personal loans in Mexico is a roughly 33% relative increase in early bad loans. That matters for three reasons. Credit is now one of MercadoLibre's biggest growth engines, with the portfolio near $14.6B in Q1 2026. Mexico is the market where it wants a banking licence and to become the country's leading digital bank, so regulators are watching asset quality. And first-time borrowers are the inclusion story: people with thin Buró de Crédito files who were approved because of MercadoLibre's own data. If those loans go bad, the inclusion bet looks reckless. I'll treat this as a credit RCA: separate mix effects from real deterioration, then look at underwriting, product, collections and the macro environment.

Step 1

Clarifying Questions (3 minutes)

• Over what period did the rate move, and is it measured on balances or on number of loans?

Why it matters: Delinquency on balances can rise because larger loans go bad, even if the count of bad loans is flat. Expected answer: Over two quarters, measured on outstanding balance. Impact on approach: I'd check both count and balance views.

• Is this a cohort view (loans originated in a given month) or the whole first-time book at a point in time?

Why it matters: A stock metric mixes old and new loans. If originations slowed, the denominator shrinks and the rate rises without any loan getting worse. Expected answer: Whole portfolio of first-time borrowers. Impact on approach: Rebuild the view by origination vintage.

• Did anything change in underwriting, limits, pricing or acquisition in the last six months?

Why it matters: Policy changes are the most common internal driver. Expected answer: Approval was expanded to more applicants without a bureau record, and maximum first-loan size was raised. Impact on approach: Expansion cohorts become the prime suspect.

• Is the increase also visible for repeat borrowers and credit card holders in Mexico, or in Brazil?

Why it matters: If everything worsens, macro is likely; if only first-time borrowers, it is policy or product. Expected answer: Repeat borrowers and Brazil are stable; Mexico card NPL slightly up. Impact on approach: Focus on first-time Mexico, but watch macro.